D700 recognised as Radeon R9 280X by 3D Mark

[iBug2]

My apologies, I'd swear I read $10K for maxed out new one.

My mistake.

In any case, your math was even more off.

The 2009 was $1,100 (current value of used 2009 Octo 2.26)
The Dual 5680s were $1,500
The Dual Titans were $2,000
The PC power supply was $45

So, I've got $4,645 in the winning machine
Vs $7,600 for the machine that LOST

Leaving $3K for a nicely appointed rMBP

There's your FREE Thunderbolt

So, faster machine and a rMBP or slower one by itself.

Tough one.

Well my math was for retail prices, obviously, not 2nd hand on the one hand and retail on the other. Your comparison is not fair.

And a maxed out nMP is 10K, but in that you get a 1TB HD and 64GB RAM. Since you are not considering those extensions in the classic Mac Pro, leave them out of the nMP as well and we get 7600$.

And your winner isn't really a winner. It's faster if you are gaming and working on CUDA apps. It's a loser everywhere else.

 

[N19h7m4r3]

Well my math was for retail prices, obviously, not 2nd hand on the one hand and retail on the other. Your comparison is not fair.

And a maxed out nMP is 10K, but in that you get a 1TB HD and 64GB RAM. Since you are not considering those extensions in the classic Mac Pro, leave them out of the nMP as well and we get 7600$.

And your winner isn't really a winner. It's faster if you are gaming and working on CUDA apps. It's a loser everywhere else.

Except that CUDA doesn't work with the new NVIDIA cards such as the B( and they only recently got OpenCL to work again) revisions of Titans 780's, and 780Ti's.

https://forums.macrumors.com/posts/18757149/

Read the thread about GK110B fix.

Download the Beta linked driver that ends with "B07".

Install on 10.9.1 and you will have FASTEST MAC GPU EVER.

(No CUDA yet but does have OpenCl)

So what you get a machine that's better at current games, and horrible at OpenCL, while lagging farther and farther behind on RAM, CPU speeds, and newer instruction sets. Hardly the "fastest" GPU in a Mac Pro.

 

[flowrider]

I have to admit: I had a little trouble deciding to post this, Dave. I've done business with you in the past and I'm quite happy with the product I've purchased. I've also recommended you and your services to other Mac Pro owners who didn't realize they had a much larger choice of GPUs for their rigs at the time.

I think you're pushing buttons and trolling, here. I actually think you're smarter than you're letting on, but let's play along for a second.

Clearly, the old Mac Pro, with a pair of (now dated) x5690s will be a CPU powerhouse. It'll be that powerhouse for applications that are properly threaded and don't spend their time constantly context switching. It'll also be a powerhouse for apps that don't need things like Intel's AVX technology; that'll never be available on the old Mac Pro, no matter how hard you try.

And yes, you can augment the power supply situation in the old beast so that you can power up 2 Titans or 2 780Ti GPUs. You'd do this assuming your application can take advantage of more than 1 GPU, and that it needs CUDA processing; note for this discussion I'm flat-out ignoring games. I don't really care about gaming on a Mac and I personally think that those that do are foolish.

With all of this assembled, you have a system that performs well today, with today's applications. It's a tactical decision. And potentially a good one.

I'm an architect by trade (and nature). Tactical thought is the antithesis of what I need to do for my day to day business. Strategic thought is more my thing, and the strategy here is: what app changes are coming in the future? Apple is clearly pushing to relegate nVidia's CUDA to a, "Awww... isn't that cute!" place. OpenCL isn't there yet, today. Apps that have been CUDA-aware for a few years are going to take time to get themselves shifted to OpenCL. But several are already on their way.

Apple wants OpenCL to win, and right now, that means AMD is the choice for GPUs. Love it or hate it, nVidia just can't (or, more accurately: won't) get their OpenCL performance up to the same level. With all that time and money invested in CUDA, why would they?

Ignoring the GPUs for a second, the CPU is something else the old Mac Pro just can't update. It's stuck at the x5690 as the top chip and will never go beyond that. As more and more applications start tapping Intel's AVX to help speed things up a bit, the old Mac Pro's CPUs will seem less and less desirable. Final Cut Pro X has been using AVX since its inception. Adobe has hinted that Premiere Pro is, as well (though I have no proof of that).

The new machine won't do as well in most benchmarks as the old Mac Pro will. And for the time being, there are a bunch of apps that might even run faster on the old Pro, assuming the owner has updated the GPU to something more modern. But it won't always be the case. Count on it.

The simple fact is that MVC is talking about TODAY not TOMORROW.

My take on what MVC is saying, is that it's not time to throw the oMP under the bus just yet. There's still a lot of life left in the old girl yet.

And for me, at least, the internal expandability of the oMP and the ease of upgrading her with off the shelf components (CPUs, USB3 cards, WiFi 802.11ac cards and etc.) or readily available flashable products (GPUs) keep the oMP viable and make it an upgraders dream.

And another thought to ponder, when the nMP reaches EOL, how upgradable do you expect it to be to be? With all it's not standard configuration components, my bet it not very.

My point, at least in the short run, two or maybe even for the next five years the oMP can be still be a very productive and fast machine.

Lou

 

[kennyman]

Except that CUDA doesn't work with the new NVIDIA cards such as the B( and they only recently got OpenCL to work again) revisions of Titans 780's, and 780Ti's.

https://forums.macrumors.com/posts/18757149/




So what you get a machine that's better at current games, and horrible at OpenCL, while lagging farther and farther behind on RAM, CPU speeds, and newer instruction sets. Hardly the "fastest" GPU in a Mac Pro.

You guys are exaggerating here. What MVC is trying to say is that you get better performance from old mac pro with new cards and it is true. How much do you benefit from AVX, AES and SSE instructions? Only Grand Central Dispatch uses it, and it does not even make a huge big difference. Same as Intel Quick Sync, which never really got implemented in rendering. I work for a design/media company and we have 2 nMP maxed out and let me tell you they are NOT GPU powerhouses or media editing dream machines that some have claimed.

We have 5-6 cMP 2010 version maxed out, some with used 5690 (yes, refurb or used OEM cpus from certified sever resellers). And those old machine with either dual AMD 7970 or Dual Titan (depending on what you using, maya, mari, adobe, render man and so on). The 2010 with dual X5690 and dual AMD 7970 will leave the 12-core nMP with D700 far behind. And this is not based on reviews, it is a fact and we have been doing extensive tests on these units for almost 1 month now. Forget about dual 7970, and old mac pro with Red Rocket card and the old AMD card will give you a stable platform with a fast pcie SAS raid card.

My in-house video editors will take the old 2010 with dual AMD cards any-day. The nMP is a good toy, but not there yet for professional. Cost-performace is the first thing you look at for delivering projects on time and being profitable. Having a nMP with Mavericks (who is not the best tuned OS that has been released btw), it is a disaster to move files over dual lan to NAS/SAN, and you cannot just invest into TB boxes that cost a fortune, pretending that...yes, this is the future and we will benefit from it.

People who requires serious OpenCL performance do not even look at nMP. And for those who do edit 4K, 5K and 8K on Mac apps, will still go for the old mac pro as you are free to choose cards, AMD, red rocket and focus on delivering project, rather than wondering why the monitor is stuck at 30Hz or why the font looks so small and spending time turning for HD DPI res. Do not get me wrong, nMP is a cool machine and even I stare at it all the time. But we are not talking about a powerhouse or ultimate machine here my friends.

 

[deconstruct60]

You guys are exaggerating here. What MVC is trying to say is that you get better performance from old mac pro with new cards and it is true. How much do you benefit from AVX, AES and SSE instructions? Only Grand Central Dispatch uses it,

The problem is not that the old Mac Pro aren't still useful to some. The core problems in this kinds of threads is the layer of 'half-truths' and FUD that gets thrown in.

GCD
"... Grand Central Dispatch

All of OS X is multicore-aware and optimized for allocating tasks across multiple cores and processors. ..."
https://developer.apple.com/technologies/mac/core.html

You can only get to these new inidividual core instructions through API designed to allocate tasks between between cores? Not.

It is cleaner to invoke a AVX/SSE computational kernel with GCD in alot of cases in apps that have to remain user responsive while they work. Necessary? not even in the slightest.

and it does not even make a huge big difference.

AVX doesn't make huge difference? There are numerous posts here on macrurmors and other places of newer AVX equipped Macs putting Mac Pro systems in the rear view mirror on FCPX encoding tasks that can leverage an AVX kernel.

Same as Intel Quick Sync, which never really got implemented in rendering.

Neither the old or new Mac Pros have Quick Sync. Purely misdirection.
If want to grumble about Apple framework/driver fumbles it is in leverage the AMD and Nvidia alternatives to Quick Sync.

People who requires serious OpenCL performance do not even look at nMP.

In some areas where individual bit integrity doesn't matter (e.g., media ) they will. Where bit integrity does (e.g., money or of real objective modeling that isn't for giggles/entertainment ) they won't.

And for those who do edit 4K, 5K and 8K on Mac apps, will still go for the old mac pro as you are free to choose cards, AMD, red rocket and focus on delivering project, rather than wondering why the monitor is stuck at 30Hz

Whether stuck a 4K at 30Hz or not has as much to do with the monitor as it does with new or old model Mac Pros.

 

[N19h7m4r3]

You guys are exaggerating here.

In what way is anything I said an exaggeration? At the moment it's clear MVC is the one doing that by making claims such as the FASTEST MAC GPU EVER.

When it's hardly the case, it's situational, and saying that AVX isn't used much is an exaggeration right there.

 

[jasonvp]

The simple fact is that MVC is talking about TODAY not TOMORROW.

And that's simply short-sighted thinking. That was my point.

...or readily available flashable products (GPUs) keep the oMP viable and make it an upgraders dream.

I addressed that in my post. At least, I think I did. Dave's focus has primarily been nVidia cards (though he has also dabbled with AMD GPUs). Hopefully, my stance is pretty clear on that, given what I posted. Today: CUDA rules. Tomorrow? ...

My point, at least in the short run, two or maybe even for the next five years the oMP can be still be a very productive and fast machine.

You're throwing time frames around that are somewhat strange. "Short run", "two" and "five" years. Which is it? Software can update significantly faster than five years. Can it be updated in two? I'm not sure. But the point is: if developers decide to jump ship from CUDA and go towards OpenCL (as Apple and others are trying to herd them towards), they'll likely get that done before five years. The oMP will be way long in the tooth by then, and it'll show. Will you have a "bought and paid for" machine at that point? Absolutely. Should you be commended for using a machine until it rusts? Totally. Will it be left way in the dust by newer Mac Pros? I'll bet money on that.

 

[spaz8]

This is definitely a situational today vs. tomorrow spin. I certainly have apps that benefit from CUDA today, but OpenCl is getting more traction constantly. It only makes sense long term to use an API all GFX cards and even integrated GPU's can leverage as a software developer.

The one point that nobody seems to mention - probably because its an unknown.. is that the nMP may not stand still either. The cMP benefits from the current upgrade ecosystem. Replacements for the PCI 3 D700, 1866 ram chips, cheaper E5 v2 cpu's will all be tech the oMP will not be able to benefit from going forward. Sure you can plug in PCI cards for USB 3, etc, but at some point PCI2 or 1.1 is gonna be a bottle neck. You also only have so many PCI slots. You end up filling them up for basic I/O like USB 3, SSD/SATA3, TB, plus your GFX card (s).

 

[kennyman]

deconstruct60

I agreed with you on some of the points that you have made. But personally, I think I am going to stay with my old mac. When the day comes, when I see that my apps are really running slow on my mac. Then I think I would decide on what to do.

 

[LongSticks]

Here is a 12 Core nMP:

http://www.3dmark.com/fs/1627960

And here is the Champion, a 12 Core cMP:

http://www.3dmark.com/3dm/2415477

No, I don't think for a second that this is the be-all, end-all test.

There are certainly PLENTY of tests where a nMP can beat mine.

But in the PC World, this test is important. And for the "Firestrike" part, the cMP is 40% faster. The other two tests are a draw.

And it declares that a 12 Core nMP D700 is faster than "93% of PCs" while my ancient dinosaur of a 12 Core somehow beat 97%. Hmmm.

But it certainly shows that despite a much slower PCIE bus and MUCH slower SSD, the "Classic" Mac Pro can be upgraded GPU-wise to equal or exceed the new one. And I don't even have 5690s yet.

And the cMP can still be upgraded further (780Ti as soon as 10.9.2 & finished Web Driver drop) whereas new one, well it's kind of done being upgraded GPU wise. Certainly for the next 6-18 months anyway.

And I guarantee that I have far less then $10K in this cMP. The nMP 12 core on the other hand....

Let's just say that you could get yourself a cMP with Dual 5680s and Dual Titans and a Power Supply....and get yourself "ThunderBolt" by adding a rMBP with the money left over compared to a nMP 12 Core.

Another quick point of view on MVC's post, it's completely Americancentric as we haven't been able to buy cMP in Europe for 2yrs. The used pool, looking at the ScrumpyMacs site a couple of weeks ago, is also drying up fast.

Secondly - over a 3yr period, what would the power consumption difference be between the MVC as specified cMP and the 12c nMP - I don't know if he can tell us the power draw, based on it working a 60-70hr week! With utility prices we have in the UK at the moment, this is factoring in to whole life budget calculations more and more.

I suppose the added heat from all those fans may give a saving on my heating bills!

 

[MacVidCards]

Except that CUDA doesn't work with the new NVIDIA cards such as the B( and they only recently got OpenCL to work again) revisions of Titans 780's, and 780Ti's.

https://forums.macrumors.com/posts/18757149/




So what you get a machine that's better at current games, and horrible at OpenCL, while lagging farther and farther behind on RAM, CPU speeds, and newer instruction sets. Hardly the "fastest" GPU in a Mac Pro.

You seem to be having reading trouble or my posts aren't clear.

The "B" cards are 100% fully functional in 10.9.1 using a piece of latest Beta driver. Full CUDA, OpenCl, etc. Run 100% just like "A" cards.

Only one to not have CUDA is the 780Ti.

And where did "Horrible" at OpenCl come from? Maybe you are getting 680 confused with 780? They are similar looking until you notice that they aren't the same. Titan/780Ti combo get 3150 on OpenCL "Sala" Benchmark that D700 twins get 3400 on, so 8-9% behind. Guess that is "horrible" now.

"Fastest GPU" is exactly that, the GPU. Like it or nor, from the day the nMP came out it was running an older pair of GPUs which can be bested by things that go in PCIE slots like those in cMP. If OpenCl power is what is needed, Dual 7970s score over 4000 in Sala, and since that is more than 10% faster than 3400, by your standards that means the D700s are "horrible".

Did anyone actually read my post? I specifically said:

"No, I don't think for a second that this is the be-all, end-all test.

There are certainly PLENTY of tests where a nMP can beat mine."

So, you can put faster, better GPUs in a cMP than anything available for a nMP.

Getting out the tar & feathers because I point this out doesn't help, it just shows that some people are a little "sensitive".

Does nMP have a disk drive that is faster than anything I can put in cMP? Yes. (until a faster PCIE one comes out anyway)

Does nMP have a CPU that uses nifty new tech and can run FCP faster clock for clock than my 5680s? Absofreakinglutely !!

Does the nMP have better, faster RAM? Yep, and with it's 4 slots it can use as much as I can with 8.

So if I can readily admit that ALL of those things are better in nMP, why is saying cMP can have better GPUs so upsetting?

A guy starts a thread showing what a nMP D700 does in 3DMark.

I post that a cMP (via PCIE upgrade) can best that by 40%. What does 4 x "horrible" equal?

EDIT: And of course when you need REAL OpenCl performance it's best to stop monkeying around with the trashcan-with-a-fan and pop some serious cards in your cMP. By removing 2 HDs I was able to slide 3rd card in. (No REAL Pro uses internal storage anymore, or so I've been told)

Fired up Luxmark again.

Whoa !!!! 4500 is more than 3500 I think.

But that's only if you need real OpenCl horsepower. Keep in mind, cost still under $7,600.

If Dual GPU OpenCl is good, isn't Triple GPU OpenCl better?

(note to self, look up "order of magnitude")

 

[MH01]

Actually, after a little research it is indeed down to things like Double Precision calculations. The cards all start out the same (as in workstation grade) but under testing if they discover faults, and if these faults can be ignored for consumer level cards then they can lock them out via software driver or even soldering the circuit board (I assume to stop soft modding). These cards end up as consumer cards or go in the trash can (no pun intended).

Correct. Though a lot of people on here will state its just drivers, they are incorrect. Pointless having tuned drivers if the hardware is not upto it.

The same goes for Xeon CPUs. The best of the parts become Xeons, ones that fail the grade become Extreme Edition CPUs etc etc.

No difference for GPUs, no magic makes the Consumer version of the GPU identical to the Firepro, its the manufacturing process and defects that define what becomes consumer and what gets chucked.

 

[MacVidCards]

Correct. Though a lot of people on here will state its just drivers, they are incorrect. Pointless having tuned drivers if the hardware is not upto it.

The same goes for Xeon CPUs. The best of the parts become Xeons, ones that fail the grade become Extreme Edition CPUs etc etc.

No difference for GPUs, no magic makes the Consumer version of the GPU identical to the Firepro, its the manufacturing process and defects that define what becomes consumer and what gets chucked.

Hear Hear !!!

Don't forget the ECC RAM, can't do important, accurate work without it.

Oh, wait.....trying to remember which machine supports GPUs with ECC RAM.

 

[antonis]

Let's not kid ourselves, it is never actually a user's decision. Since Apple has chosen to move forward with the nMP, this is the way to go, convenient or not. That is, for anyone who wants to remain within Apple's ecosystem.

Apple doesn't like to look back, so trying to keep alive a discontinued model it is a - noble but - lost game.

For any professional out there who opposes with the ideas behind the nMP, the only serious alternative is to start looking outside Apple's ecosystem.

 

[nathan43082]

Whoa !!!! 4500 is more than 3500 I think.

I know that down-clocking for the sake of aiding cooling is one of the reasons the nMP does not hit those higher benchmarks. Engineering is all about trade-offs, I suppose, and one can choose the nMP because they like the choices Apple made or they can opt for some other solution. I am curious as to the power draw on that old Mac Pro and how loud those cards are when the thing is sitting near where you work.